EEAT Mechanics

Freshness twiddler: a second version and a setting for Human Voices

Beyond the FreshnessTwiddler known from the leak, the schema list shows a FreshnessTwiddler2 with its own settings for HuVo, short for Human Voices. What the second version changes and how HuVo is treated differently is not known.

By Thomas Wawra· Published · Version 1.0· Systems referenced: FreshnessTwiddler / RealTimeBoost, Date triangulation

01What the schemas add

The query-dependent freshness system has its own deep dive on this site, /deep-dive/freshness-twiddler. It covers Query Deserves Freshness, the patent on update scoring and staleness, and what that means for updates. This article does not repeat it. It covers what the schema list adds.

There are four objects. FreshnessTwiddlerDocAnnotations holds information attached to each document, FreshnessTwiddlerQueryAnnotations information attached to the query. Two more belong to FreshnessTwiddler2: FreshnessTwiddler2Params, the settings sent with the query, and FreshnessTwiddler2HuVoParams, a separate set of settings for HuVo.

RESONEO expands HuVo as Human Voices, a label it applies to forum posts, social media and creator material. If that reading is right, freshness is tuned differently for such content than for ordinary pages. In which direction, and for which queries, the schema names do not say. Nor do they say whether FreshnessTwiddler2 replaced the first version or runs beside it.

Source per claim (3)
C
The schema list shows four freshness objects: FreshnessTwiddlerDocAnnotations, FreshnessTwiddlerQueryAnnotations, FreshnessTwiddler2Params and FreshnessTwiddler2HuVoParams.[1]
C
RESONEO reads HuVo as Human Voices, meaning forum, social and creator content, and concludes that such content gets its own freshness setting.[2]
C
Whether FreshnessTwiddler2 replaced the first version or runs alongside it, and how the HuVo settings differ, cannot be read from the schema names.[3]

02What the leak says

The leaked API documentation names the twiddler in one place we have checked. In PerDocData, the large per-document record, a field with date-related information carries the note that it is used in FreshnessTwiddler, with the example that a page is old judged by its date annotations.

RESONEO goes further and says the twiddler reads both a page's stated dates and a date Google estimates for its content, derived from the page text, inbound anchor text and associated pages. That is RESONEO's summary; we have not verified this wording in the leak.

Source per claim (2)
B
PerDocData describes a field holding "…dates-related info (e.g. page is old based on its date annotations). Used in FreshnessTwiddler."[4]
C
According to RESONEO, FreshnessTwiddler combines a page's dates with an estimated content date derived from page text, inbound anchors and associated pages.[2]

03What Google says publicly

Google confirms the principle, not the twiddler. Its guide to ranking systems mentions several "query deserves freshness" systems for queries where fresher content is expected. Its documentation on byline dates says that Google looks at several date factors, because any single one can be wrong.

That second point is why the freshness twiddler depends on date triangulation (system 25): the twiddler can only judge freshness as well as the date it is given. Neither Google page mentions HuVo, a second version, or different freshness rules for forum and social content.

Source per claim (3)
O
"We have various "query deserves freshness" systems designed to show fresher content for queries where it would be expected."[5]
O
"Google doesn't depend on a single date factor because all factors can be prone to issues."[6]
C
No Google page we found mentions HuVo or a separate freshness treatment for forum, social or creator content.[3]

04E-E-A-T and what a site owner can do

Freshness touches E-E-A-T only through trust, and only indirectly: an outdated answer to a time-sensitive question lets the reader down. Otherwise it is a timing rule. What a site owner controls is the date signal. Google asks for a visible, labelled date and for dates that agree between the page and its structured data. A faked date gains nothing if the dates do not agree, which is the point of triangulation.

Source per claim (2)
O
Google recommends a prominent, labelled date on the page and advises: "Make your dates and times consistent," between the visible and the structured values.[6]
C
Freshness bears on E-E-A-T only indirectly, through trust; the system card rates E, Expertise and A as none.[3]

05References

This Deep Dive is Layer 2 content. It interprets the material and always points back to Layer 1, the reference layer. Every claim is mapped to a source, labeled by type: [A] Sworn testimony · [B] Verbatim from the leak · [O] Documented by Google · [W] Peer-reviewed study · [S] Checked on the page · [C] Our assessment · [P] Patent.

© Thomas Wawra · Senior SEO Manager · wetter.com — a Funke Digital company